This repo is for those who would like to start learning machine learning algorithms and the machine pipeline.
- Titanic Survival Classification
[Project Background] This is for survival prediction on Titanic Data. The model is built to predict which passenger survived on Titanic shipwreck.
[Dataset]
[Model]
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
dense (Dense) (None, 32) 1792
_________________________________________________________________
dropout (Dropout) (None, 32) 0
_________________________________________________________________
batch_normalization (BatchNo (None, 32) 128
_________________________________________________________________
dense_1 (Dense) (None, 16) 528
_________________________________________________________________
dropout_1 (Dropout) (None, 16) 0
_________________________________________________________________
batch_normalization_1 (Batch (None, 16) 64
_________________________________________________________________
dense_2 (Dense) (None, 5) 85
_________________________________________________________________
batch_normalization_2 (Batch (None, 5) 20
_________________________________________________________________
dense_3 (Dense) (None, 1) 6
=================================================================
Total params: 2,623
Trainable params: 2,517
Non-trainable params: 106
[Evaluation]
Classification Report:
precision recall f1-score support
0.0 0.78 0.93 0.85 157
1.0 0.86 0.63 0.73 111
accuracy 0.81 268
macro avg 0.82 0.78 0.79 268
weighted avg 0.82 0.81 0.80 268
- Pandas, numpy, os,
- matplotlib, seaborn,
- tensorflow, scikit-learn, keras